[cleaner] Add option to skip cleaning files

A new option --skip-clean-files allows cleaner to skip cleaning files
where the user is certain no sensitive information is present.

The option supports globs / wildcards.

Relevant: #3469
Closes: #3520

Signed-off-by: Pavel Moravec <pmoravec@redhat.com>

man/en/sos-clean.1

@@ -5,6 +5,7 @@ sos clean - Obfuscate sensitive data from one or more sosreports
 .B sos clean TARGET [options]
     [\-\-domains]
     [\-\-disable-parsers]
+    [\-\-skip-clean-files]
     [\-\-keywords]
     [\-\-keyword-file]
     [\-\-map-file]
@@ -63,6 +64,15 @@ trust in the party/parties that may handle the generated report.
 Valid values for this option are currently: \fBhostname\fR, \fBip\fR, \fBipv6\fR,
 \fBmac\fR, \fBkeyword\fR, and \fBusername\fR.
 .TP
+.B \-\-skip-clean-files FILES
+Provide a comma-delimited list of files inside an archive, that cleaner should skip in cleaning.
+
+Globs like asterisk are supported, so \fBsos_commands/host/hostname*\fR will match all three
+usual filenames in that directory (\fBhostname\fR, \fBhostnamectl_status\fR and \fBhostname_-f\fR).
+
+Use this option with caution, only when being certain the given files do not contain any sensitive
+information.
+.TP
 .B \-\-keywords KEYWORDS
 Provide a comma-delimited list of keywords to scrub in addition to the default parsers.
 

sos/cleaner/__init__.py

@@ -15,6 +15,7 @@
 import shutil
 import sos.cleaner.preppers
 import tempfile
+import fnmatch
 
 from concurrent.futures import ThreadPoolExecutor
 from datetime import datetime
@@ -81,6 +82,7 @@ class SoSCleaner(SoSComponent):
         'archive_type': 'auto',
         'domains': [],
         'disable_parsers': [],
+        'skip_clean_files': [],
         'jobs': 4,
         'keywords': [],
         'keyword_file': None,
@@ -116,7 +118,7 @@ def __init__(self, parser=None, args=None, cmdline=None, in_place=False,
             # when obfuscating a SoSCollector run during archive extraction
             os.makedirs(os.path.join(self.tmpdir, 'cleaner'), exist_ok=True)
 
-        self.validate_parser_values()
+        self.review_parser_values()
 
         self.cleaner_mapping = self.load_map_file()
         os.umask(0o77)
@@ -125,13 +127,14 @@ def __init__(self, parser=None, args=None, cmdline=None, in_place=False,
 
         self.cleaner_md = self.manifest.components.add_section('cleaner')
 
+        skip_clean_files = self.opts.skip_clean_files
         self.parsers = [
-            SoSHostnameParser(self.cleaner_mapping),
-            SoSIPParser(self.cleaner_mapping),
-            SoSIPv6Parser(self.cleaner_mapping),
-            SoSMacParser(self.cleaner_mapping),
-            SoSKeywordParser(self.cleaner_mapping),
-            SoSUsernameParser(self.cleaner_mapping)
+            SoSHostnameParser(self.cleaner_mapping, skip_clean_files),
+            SoSIPParser(self.cleaner_mapping, skip_clean_files),
+            SoSIPv6Parser(self.cleaner_mapping, skip_clean_files),
+            SoSMacParser(self.cleaner_mapping, skip_clean_files),
+            SoSKeywordParser(self.cleaner_mapping, skip_clean_files),
+            SoSUsernameParser(self.cleaner_mapping, skip_clean_files)
         ]
 
         for _parser in self.opts.disable_parsers:
@@ -262,6 +265,10 @@ def add_parser_options(cls, parser):
                                default=[], dest='disable_parsers',
                                help=('Disable specific parsers, so that those '
                                      'elements are not obfuscated'))
+        clean_grp.add_argument('--skip-clean-files', action='extend',
+                               default=[], dest='skip_clean_files',
+                               help=('List of files to skip/ignore during '
+                                     'cleaning. Asterisks are supported.'))
         clean_grp.add_argument('-j', '--jobs', default=4, type=int,
                                help='Number of concurrent archives to clean')
         clean_grp.add_argument('--keywords', action='extend', default=[],
@@ -323,17 +330,20 @@ def inspect_target_archive(self):
         if self.nested_archive:
             self.nested_archive.ui_name = self.nested_archive.description
 
-    def validate_parser_values(self):
-        """Check any values passed to the parsers via the commandline, e.g.
-        the --domains option, to ensure that they are valid for the parser in
-        question.
+    def review_parser_values(self):
+        """Check any values passed to the parsers via the commandline:
+        - For the --domains option, ensure that they are valid for the parser
+          in question.
+        - Convert --skip-clean-files from globs to regular expressions.
         """
         for _dom in self.opts.domains:
             if len(_dom.split('.')) < 2:
                 raise Exception(
                     f"Invalid value '{_dom}' given: --domains values must be "
                     "actual domains"
                 )
+        self.opts.skip_clean_files = [fnmatch.translate(p) for p in
+                                      self.opts.skip_clean_files]
 
     def execute(self):
         """SoSCleaner will begin by inspecting the TARGET option to determine

sos/cleaner/archives/__init__.py

@@ -50,7 +50,6 @@ class SoSObfuscationArchive():
     type_name = 'undetermined'
     description = 'undetermined'
     is_nested = False
-    skip_files = []
     prep_files = {}
 
     def __init__(self, archive_path, tmpdir):

sos/cleaner/parsers/__init__.py

@@ -42,22 +42,24 @@ class SoSCleanerParser():
     name = 'Undefined Parser'
     regex_patterns = []
     skip_line_patterns = []
-    skip_files = []
+    parser_skip_files = []  # list of skip files relevant to a parser
+    skip_clean_files = []   # list of global skip files from cmdline arguments
     map_file_key = 'unset'
     compile_regexes = True
 
-    def __init__(self, config={}):
+    def __init__(self, config={}, skip_clean_files=[]):
         if self.map_file_key in config:
             self.mapping.conf_update(config[self.map_file_key])
+        self.skip_clean_files = skip_clean_files
         self._generate_skip_regexes()
 
     def _generate_skip_regexes(self):
-        """Generate the regexes for the parser's configured `skip_files`,
-        so that we don't regenerate them on every file being examined for if
-        the parser should skip a given file.
+        """Generate the regexes for the parser's configured parser_skip_files
+        or global skip_clean_files, so that we don't regenerate them on every
+        file being examined for if the parser should skip a given file.
         """
         self.skip_patterns = []
-        for p in self.skip_files:
+        for p in self.parser_skip_files + self.skip_clean_files:
             self.skip_patterns.append(re.compile(p))
 
     def generate_item_regexes(self):

sos/cleaner/parsers/hostname_parser.py

@@ -21,9 +21,9 @@ class SoSHostnameParser(SoSCleanerParser):
         r'(((\b|_)[a-zA-Z0-9-\.]{1,200}\.[a-zA-Z]{1,63}(\b|_)))'
     ]
 
-    def __init__(self, config):
+    def __init__(self, config, skip_clean_files=[]):
         self.mapping = SoSHostnameMap()
-        super(SoSHostnameParser, self).__init__(config)
+        super(SoSHostnameParser, self).__init__(config, skip_clean_files)
 
     def parse_line(self, line):
         """This will be called for every line in every file we process, so that

sos/cleaner/parsers/ip_parser.py

@@ -25,7 +25,7 @@ class SoSIPParser(SoSCleanerParser):
         r'.*dnf\[.*\]:'
     ]
 
-    skip_files = [
+    parser_skip_files = [
         # skip these as version numbers will frequently look like IP addresses
         # when using regex matching
         'installed-debs',
@@ -44,6 +44,6 @@ class SoSIPParser(SoSCleanerParser):
     map_file_key = 'ip_map'
     compile_regexes = False
 
-    def __init__(self, config):
+    def __init__(self, config, skip_clean_files=[]):
         self.mapping = SoSIPMap()
-        super(SoSIPParser, self).__init__(config)
+        super(SoSIPParser, self).__init__(config, skip_clean_files)

sos/cleaner/parsers/ipv6_parser.py

@@ -29,15 +29,15 @@ class SoSIPv6Parser(SoSCleanerParser):
         r"(([0-9a-f]{1,4}(:[0-9a-f]{0,4}){0,5}))([^.])::(([0-9a-f]{1,4}"
         r"(:[0-9a-f]{1,4}){0,5})?))(/\d{1,3})?(?![:\\a-z0-9])"
     ]
-    skip_files = [
+    parser_skip_files = [
         'etc/dnsmasq.conf.*',
         '.*modinfo.*',
     ]
     compile_regexes = False
 
-    def __init__(self, config):
+    def __init__(self, config, skip_clean_files=[]):
         self.mapping = SoSIPv6Map()
-        super(SoSIPv6Parser, self).__init__(config)
+        super(SoSIPv6Parser, self).__init__(config, skip_clean_files)
 
     def get_map_contents(self):
         """Structure the dataset contents properly so that they can be reloaded

sos/cleaner/parsers/keyword_parser.py

@@ -20,9 +20,9 @@ class SoSKeywordParser(SoSCleanerParser):
     name = 'Keyword Parser'
     map_file_key = 'keyword_map'
 
-    def __init__(self, config):
+    def __init__(self, config, skip_clean_files=[]):
         self.mapping = SoSKeywordMap()
-        super(SoSKeywordParser, self).__init__(config)
+        super(SoSKeywordParser, self).__init__(config, skip_clean_files)
 
     def _parse_line(self, line):
         return line, 0

sos/cleaner/parsers/mac_parser.py

@@ -43,15 +43,15 @@ class SoSMacParser(SoSCleanerParser):
         '53:4f:53',
         '534f:53'
     )
-    skip_files = [
+    parser_skip_files = [
         'sos_commands/.*/modinfo.*'
     ]
     map_file_key = 'mac_map'
     compile_regexes = False
 
-    def __init__(self, config):
+    def __init__(self, config, skip_clean_files=[]):
         self.mapping = SoSMacMap()
-        super(SoSMacParser, self).__init__(config)
+        super(SoSMacParser, self).__init__(config, skip_clean_files)
 
     def reduce_mac_match(self, match):
         """Strips away leading and trailing non-alphanum characters from any

sos/cleaner/parsers/username_parser.py

@@ -26,9 +26,9 @@ class SoSUsernameParser(SoSCleanerParser):
     map_file_key = 'username_map'
     regex_patterns = []
 
-    def __init__(self, config):
+    def __init__(self, config, skip_clean_files=[]):
         self.mapping = SoSUsernameMap()
-        super(SoSUsernameParser, self).__init__(config)
+        super(SoSUsernameParser, self).__init__(config, skip_clean_files)
 
     def _parse_line(self, line):
         return line, 0

sos/collector/__init__.py

@@ -87,6 +87,7 @@ class SoSCollector(SoSComponent):
         'group': None,
         'image': '',
         'force_pull_image': True,
+        'skip_clean_files': [],
         'jobs': 4,
         'journal_size': 0,
         'keywords': [],
@@ -483,6 +484,10 @@ def add_parser_options(cls, parser):
                                  default=[], dest='disable_parsers',
                                  help=('Disable specific parsers, so that '
                                        'those elements are not obfuscated'))
+        cleaner_grp.add_argument('--skip-clean-files', action='extend',
+                                 default=[], dest='skip_clean_files',
+                                 help=('List of files to skip/ignore during '
+                                       'cleaning. Asterisks are supported.'))
         cleaner_grp.add_argument('--keywords', action='extend', default=[],
                                  dest='keywords',
                                  help='List of keywords to obfuscate')

sos/report/__init__.py

@@ -88,6 +88,7 @@ class SoSReport(SoSComponent):
         'desc': '',
         'domains': [],
         'disable_parsers': [],
+        'skip_clean_files': [],
         'dry_run': False,
         'estimate_only': False,
         'experimental': False,
@@ -358,6 +359,10 @@ def add_parser_options(cls, parser):
                                  default=[], dest='disable_parsers',
                                  help=('Disable specific parsers, so that '
                                        'those elements are not obfuscated'))
+        cleaner_grp.add_argument('--skip-clean-files', action='extend',
+                                 default=[], dest='skip_clean_files',
+                                 help=('List of files to skip/ignore during '
+                                       'cleaning. Asterisks are supported.'))
         cleaner_grp.add_argument('--keywords', action='extend', default=[],
                                  dest='keywords',
                                  help='List of keywords to obfuscate')